Applications of 2-D-Nmr Maps and Geometric Pore Scale Modeling for Petrophysical Evaluation of a Gas Well

Pedro A. Romero¹, Guillermo Azpiroz², Mikhail Gladkikh³, and Gustavo Magenta^1
1Baker Atlas Geoscience, Baker Hughes, Buenos Aires, Argentina
²Distal Areas, Argentina, Repsol-YPF, Comodoro Rivadavia, Argentina
³Inteq, Baker Hughes, Houston, TX

This paper presents the petrophysical evaluation of a gas reservoir from the Gulf of San Jorge basin, Argentina, by means of 2D-NMR maps and a geometric pore scale modeling of the NMR response of the wetting phase. The reservoir contains gas and irreducible water saturation at the top with a sharp transition into the water zone.

The well has been logged with conventional tools as Gamma Ray, Neutron, Density and also the MRExplorerSM (MREXSM) to acquire the Nuclear Magnetic Resonance data for determining the irreducible water saturation. Data from Special Core Analysis (SCAL), Scanning Electron Microscopy (SEM), thin sections and capillary pressure tests have also been acquired. The core-log evaluation shows a very good agreement between the laboratory and log field data, especially in terms of irreducible water saturation, porosity and permeability, which was also modeled using the Timur-Coates equation for purposes of field delivery. The 2D-NMR maps as T1-T2apparent and Diffusivity-T2intrinsic from both, the hydrocarbon and water zone, lead to characterize the clay-bound and capillary-bound water, and the under-called porosity due to the low hydrogen index of the gas.

We use a pore scale model to predict, in a good agreement with experimental data, the porosity, permeability, irreducible water saturation and T2 responses of the wetting phase, including the pendular water effect. A key feature of this model is that it is geometrically determined or precisely defined based on the knowledge of geometry and, hence, the morphology of the pore space at the grain scale. Unlike many other approaches to pore-level modeling, our approach introduces no adjustable parameters and can be used to produce quantitative, a priori predictions of the rock macroscopic behavior.

AAPG Search and Discovery Article #90078©2008 AAPG Annual Convention, San Antonio, Texas